pTCP
 
Site Index
Home
Vision
People
Publications
Projects
Current
Atra
Smarta
pTCP / R2CP
Thinck
Garuda
SCT
CtS
WSAN
Past
Sphinx
iQ
eCo
GTCP
MPFD
Sponsors
Links
Contact


Overview |  Results |  Publications |  Software |  People |  References


Overview:

The TCP transport layer protocol is designed for connections that traverse a single path between the sender and receiver. However, there are several environments in which multiple paths can be used by a connection simultaneously. In this project we consider the problem of supporting striped connections that operate over multiple paths. We propose an end-to-end transport layer protocol called pTCP that allows connections to enjoy the aggregate bandwidths offered by the multiple paths, irrespective of the individual characteristics of the paths. We show that pTCP can have a varied range of applications through instantiations in three different environments:

  • multi-homed bandwidth aggregation,
  • weighted rate differentiation, and
  • end-systems based network striping.

Results / Status:


Multi-homed Bandwidth Aggregation

Pipe Number Blackout
Figure 1Figure 2

In Figures 1 and 2, we present the simulation results showing the performance of pTCP when used on a multi-homed mobile host for bandwidth aggregation. We compare the performance of pTCP against that of application striping opening multiple TCP sockets: the unaware application stripes in a round-robin fashion, while the smart application stripes based on a coarse-grained bandwidth estimation.

Weighted Rate Differentiation

Throughput Fairness
Figure 3Figure 4

In Figures 3 and 4, we show the performance of pTCP when used over a single path for achieving weighted rate differentiation. We compare the performance of pTCP against that of weighted AIMD and application striping. For a desired weight of w, the weighted AIMD approach modifies the congestion control parameter of default TCP from (a, b) to (aw,b/w) to emulate the performance of w aggregate TCP flows. The application striping approach simply opens w default TCP sockets, where the sending application performs round-robin striping, and the receiving application performs in-sequence read, to achieve the desired rate differentiation.

Publications & Presentations:


Software Downloads:


People:

  • Aravind Velayutham (Student)
  • Hung-Yun Hsieh (Alumnus)
  • Kyu-Han Kim (Alumnus)
  • Raghupathy Sivakumar (Professor)

References & Related Work:


Network Striping

  • Striping
    H. Adiseshu, G. Parulkar, and G. Varghese, ``A Reliable and Scalable Striping Protocol,'' Proceedings of ACM SIGCOMM, Palo Alto, CA USA, Aug. 1996.
  • Inverse Multiplexing
    J. Duncanson, ``Inverse Multiplexing,'' IEEE Communications Magazine, vol. 32, no. 4, pp. 34-41, Apr. 1994.

Bandwidth Aggregation for Mobile Hosts

  • Adaptive Inverse Multiplexing
    A. Snoeren, ``Adaptive inverse multiplexing for wide-area wireless networks,'' Proceedings of IEEE GLOBECOM, Rio de Janeireo, Brazil, Dec. 1999.
  • RMTP
    L. Magalhaes and R. Kravet, ``Transport Level Mechanisms for Bandwidth Aggregation on Mobile Hosts,'' Proceedings of IEEE ICNP 2001, Riverside, CA USA, Nov. 2001.
  • SCTP
    R. Stewart et al, ``Stream Control Transmission Protocol,'' IETF RFC 2960, Oct. 2000.

End-to-End Weighted Service Differentiation

  • MulTCP
    J. Crowcroft and P. Oechslin, ``Differentiated End-to-End Internet Services Using a Weighted Proportional Fair Sharing TCP,'' ACM Computer Communication Review, vol. 28, no. 3, pp. 53-69, July 1998.
  • TCP-LASD
    T. Nandagopal, K.-W. Lee, J.-R. Li, and V. Bharghavan, ``Scalable Service Differentiation Using Purely End-to-End Mechanisms: Features and Limitations,'' Proceedings of IWQoS, Pittsburgh, PA USA, June 2000.